figure 6 figure 7
I'm giving a few examples just to show that this is certainly not something to do with just software. My personal favored example is science, with which I'm familiar. In Europe, the scientific thought model is considered to have start in the 1600s and 1700s, and it replaced a different thought model completely. It took a long time. Social changes are really never fast, but science basically replaced the notion of alchemy and replaced it with something that is very much based on the ideas of evolution. Combination of existing ideas to create new ideas is how the big scientific discoveries come about, and the small scientific discoveries, the ones that may not win Nobel Prizes, the small incremental improvements, are equally important. Competition and selection, the notion of having to be able to disprove a theory and using that as a selection criterion for goodness or badness, is very much a part of the scientific thought model. The model it displaced was the model of alchemy, where you had closely guarded secrets and you tried to keep as much knowledge to yourself because you used that knowledge to improve your personal life. That was intellectual property in the Middle Ages in Europe. I'm going to give another contrast, religion. I'm not saying that science and religion are, in any way, enemies. I'm only trying to show that they have very different thought models. While science has a thought model that tries to encourage evolution, religions, most of them at least, tend to consider both mutation and selection harmful. Religions often do not want change in their basic knowledge. They have a body of knowledge that is considered sacred and the sacredness is a barrier to change. You don't have intellectual property but you have another barrier to evolution. Figure 6 Every software developer is very aware of one of the big notions in evolution, namely incremental improvements. Software, as we know it today, could not exist without incremental improvements. It's very expensive to create software so you have to take existing pieces of software and slowly improve them. Everybody knows this. But other parts of evolution are actually things that traditional software development, or at least commercial software development, has not been very happy with -- especially competition. Competition is bad. Competition is something most companies would rather not have because it's very expensive. It is expensive both within the product itself, where you need to have competing lines of development to see which one is actually the best, as well as between competing lines of development. It's also very tough to have competition as part of the ecosystems, where you have competing companies trying to go for the same niche, because competing companies all have a very high barrier to entry in terms of the initial sunk cost of creating a product. And, as we all know, software has inherent networking effects, which means that once you get dominant in a niche it's very hard to displace you. We in the software industry also have the problem of trying to limit the recombination aspect of evolution, again through intellectual property, and this is used as a barrier to entry in traditional commercial software development. Figure 7